3434 Ecology, 83(12), 2002, pp. 3434–3448 q 2002 by the Ecological Society of America DO ALTERNATE STABLE COMMUNITY STATES EXIST IN THE GULF OF MAINE ROCKY INTERTIDAL ZONE? MARK D. BERTNESS, 1 GEOFFREY C. TRUSSELL, 2 PATRICK J. EWANCHUK, AND BRIAN R. SILLIMAN Department of Ecology and Evolutionary Biology, Box G-W, Brown University, Providence, Rhode Island 02912 USA Abstract. It has recently been hypothesized that intertidal mussel beds and seaweed canopies in the Gulf of Maine are alternate community stable states or disturbance patch mosaics dominated by either seaweed or mussel communities. The community that occurs in a given site is proposed to be stochastic and dependent on the size of the original disturbance and subsequent recruit availability. Large disturbances are postulated to be dominated by mussel beds and barnacles with widely dispersed larvae, whereas smaller disturbances are dominated by seaweeds, with limited dispersal. Positive feedbacks are proposed to maintain these two communities. We tested this hypothesis in a tidal estuary in central Maine. At eight mussel bed and eight seaweed canopy sites, we created 9-m 2 and 1-m 2 clearings and an unmanipulated control area, and in each plot established control, caged, and cage control quadrats. After three years of monitoring, our results do not support the alternate stable state hypothesis. Instead, they suggest that the occurrence of mussel beds and seaweed canopies is highly deterministic. Seaweed canopies dominate habitats with relatively little water movement, whereas mussel beds dominate habitats with high flows; and largely independent of dis- turbance size, mussel beds and seaweed canopies rapidly returned to their original com- munity type, but only in the absence of consumers (crabs and snails). With consumers present, neither community showed significant signs of recovery, even after three years. In the presence of consumers, community recovery appears to be dependent on cracks and crevices providing refuges from consumers to seaweed and mussel recruits. The idea that natural communities may represent stochastically determined alternate stable states has important implications for understanding and managing natural ecosystems, but the very existence of alternate stable states in nature has been difficult to establish. Our results suggest that intertidal seaweed canopies and mussel beds in tidal rivers in the Gulf of Maine are highly deterministic alternative community states under consumer control. More generally, since all proposed examples of alternate community stable states are based on indirect, inferential evidence, our results imply that stochastically determined alternate community stable states might be an interesting theoretical idea without a definitive em- pirical example. Key words: alternative community stable state; Ascophyllum nodosum; Carcinus maenus; con- sumer control of community pattern; disturbance theory; Fucus spp.; Littorina littorea; predation; Mytilus edulis; rocky intertidal; secondary succession; Semibalanus balanoides. INTRODUCTION The idea that natural communities may have alter- nate stable states, where more than one distinctive and persistent type of community can occur in a given hab- itat, is an important concept. If alternate community stable states are common, they would be of consider- able conceptual interest to ecologists and important for the management and conservation of natural ecosys- tems. Alternate stable states have been suggested to play an important role in structuring a wide variety of marine and terrestrial communities, including subtidal rocky bottoms (Simenstad et al. 1978, Barkai and Branch 1988, Johnson and Mann 1988), tropical coral Manuscript received 25 October 2001; revised 16 May 2002; accepted 16 May 2002; final version received 24 May 2002. 1 E-mail: mark bertness@brown.edu 2 Present address: Marine Science Center, Northeastern University, East Point Nahant, Massachusetts 01908 USA. reefs (Knowlton 1992, Hughes 1994), tropical grass- lands (Dublin et al. 1990), and temperate forests (Pe- traitis and Latham 1999). Unfortunately, there have not been rigorous experimental tests of the actual existence of alternate community stable states in any community (Connell and Sousa 1983, Petraitis and Latham 1999). Because of our good general understanding of tem- perate rocky shore communities, they are an ideal mod- el system for a critical examination of alternate com- munity stable state theory. Rocky intertidal communities have long been im- portant model systems for the exploration of the bio- logical and physical forces that generate patterns in natural communities. Early studies clarified the roles of consumers (Paine 1966, Lubchenco 1978), compe- tition for resources (Connell 1961, Buss and Jackson 1979), and physical disturbance (Dayton 1971, Sousa 1979) in the generation of community patterns within sites. More recent studies have addressed the causes of